Mining for Product Classification Structures for Internet-Based Product Searching

1. A computer-implemented method for classifying search results responsive to a search query, comprising the steps of: maintaining, by a computer, a product database that associates each of a plurality of key words with a respective one or more structural parameters, each structural parameter comprising one of a product type, a brand name, a manufacturer, a product attribute, a product retailer, a product line, and a product discount; receiving, by the computer, a search query; identifying, by the computer, a match between one of a plurality of key words and at least a portion of the search query; in response to determining that at least one of the key words matches at least a portion of the search query, associating, by the computer, each structural parameter of each matching key word with the matching portion of the search query; verifying, by the computer, a contextual similarity between the matching portion of the search query and the structural parameter associated with the identified key words, the contextual similarity verification comprising parsing web content using a pre-determined contextual phrase, the contextual phrase comprising the structural parameter associated with each matching key word; in response to determining the contextual similarity between the one or more matching key words and the identified structural parameter, extracting, by the computer, product names associated with the matching key words in the product database and associated product URL information; and displaying, by the computer, a search result comprising the extracted product names and product URL information arranged by the identified structural parameters.

2. The method of claim 1 , wherein the step of maintaining the product database comprises the step of obtaining the key words from at least one of product search feeds for known products, web clicks on search query results, product queries in product logs, website site-maps, product categories on manufacturer websites, and product categories in manufacturer brochures.

3. The method of claim 1 , wherein at least a portion of the key words are generated based on search queries.

4. The method of claim 1 , further comprising updating the product database based on the determining step by adding a new key word, which is based on the at least a portion of the search query, to the product database in response to determining that none of the key words match the search query.

5. A computer-implemented method for classifying search results responsive to a search query, comprising the steps of: receiving, by a computer, a search query; identifying, by the computer, at least one of a plurality of key words that matches at least a portion of the search query, each matched key word having associated therewith at least one structural parameter; associating, by the computer, each structural parameter of each matching key word with the matching portion of the search query; verifying, by the computer, a contextual similarity between the matching portion of the search query and the structural parameter associated with the identified key words, the contextual similarity verification comprising parsing web content using a pre-determined contextual phrase, the contextual phrase comprising the structural parameter associated with the matching key word; and generating, by the computer, a search results webpage comprising search suggestions that are responsive to the search query and grouped by the structural parameters.

6. The method of claim 5 , wherein one of the structural parameters comprises a brand name category.

7. The method of claim 5 , wherein one of the structural parameters comprises a product attribute category.

8. The method of claim 5 , wherein one of the structural parameters comprises a product retailer category.

9. The method of claim 5 , wherein one of the structural parameters comprises a product category.

10. The method of claim 5 , wherein one of the structural parameters comprises a product line category.

11. The method of claim 5 , wherein one of the structural parameters comprises a product discount category.

12. The method of claim 5 , wherein the matching portion of the search query comprises information regarding a type of product, and wherein the search results webpage further comprises a list of items corresponding to the type of product, the list being arranged by popularity of the items.

13. The method of claim 5 , wherein the matching portion of the search query comprises at least one of a word component and a phrase component, wherein the phrase component comprises a combination of at least two words.

14. The method of claim 5 , wherein the step of associating each structural parameter with the matching portion of the search query is automatically implemented by a computer-coded automated learning software.

15. The method of claim 5 , wherein the identifying step comprises the step of evaluating a semantic comparison between the key words and at least a portion of the search query.

16. The method of claim 5 , further comprising the step of obtaining the plurality of existing key words from at least one of product search feeds, web clicks on search query results, product queries logged in product logs, website site-maps, product categories on manufacturer websites, and product categories in manufacturer brochures.

17. The method of claim 5 , wherein the existing key words comprise a product brand name, and wherein the identifying step further comprises the step of identifying a match between the product brand name and the at least a portion of the search query only if the computer verifies the match based on a number of clicks recorded in a product search log for the product brand name.

18. The method of claim 5 , wherein the contextual phrase comprises the structural parameter associated with the identified key words.

19. The method of claim 5 , wherein the web content comprises at least one of a social networking website, a blog, a forum website, a product website, and a review website.

20. A computer-implemented method for mining classification structures for search results from textual content, comprising the steps of: receiving, by a computer, textual content; identifying, by the computer, a match between one of a plurality of key words and at least a portion of the received textual content, the matching key word having associated therewith at least one structural parameter that classifies the key word; verifying a contextual similarity between the match and the structural parameter associated with each matching key word, the contextual similarity verification comprising parsing web content using a contextual phrase, the contextual phrase comprising the structural parameter associated with each matching key word; in response to verifying the match, classifying, by the computer, the at least a portion of the received textual content with the structural parameter associated with the matching key word; and storing, by the computer, in a product database, the at least a portion of the received textual content with its associated structural parameter.

21. The method of claim 20 , wherein the classifying step is automatically implemented by a computer-coded automated learning software.

22. The method of claim 20 , wherein the identifying step comprises the step of evaluating a semantic comparison between the plurality of existing key words and the at least a portion of the received textual content.

23. The method of claim 20 , wherein the plurality of existing key words are retrieved from at least one of product search feeds, web clicks on search query results, product queries logged in product logs, website site-maps, product categories on manufacturer websites, and product categories in manufacturer brochures.

24. The method of claim 20 , wherein the contextual phrase comprises the structural parameter associated with each matching key word.

25. The method of claim 20 , wherein the web content comprises at least one of a social networking website, a blog, a forum website, a product website, and a review website.

26. The method of claim 20 , wherein the pre-determined contextual phrase defines the at least a portion of the received textual content as being at least one of a brand name, a manufacturer, a retailer name, a store name, a place name, and a color.

27. The method of claim 20 , further comprising: receiving, by the computer, a search query; determining, by the computer, whether at least one of the key words in the product database matches at least a portion of the search query, each key word having one or more structural parameters associated therewith; in response to determining that at least one of the key words matches at least a portion of the search query, verifying, by the computer, a contextual similarity between the matching portion of the search query and the identified structural parameter, the contextual similarity verification comprising parsing web content using a pre-determined contextual phrase; and updating, by the computer, the product database based on the verifiying step.

28. The method of claim 27 , wherein the step of updating the product database based on the determining step comprises the step of increasing a confidence score of an association between each matched key word and the at least a portion of the search query in response to determining that at least one of the key words matches at least a portion of the search query.